# Cross-Sectional Dispersion Trading (JP Growth Stocks)

- **Slug**: `dispersion-jp-growth`
- **Verdict**: rejected (regime_dependent)
- **Asset class**: equity
- **Categories**: dispersion, regime, timing
- **Source**: qbuntu-analysis (computed asof 2026-07-31T01:32:21)
- **Reproducibility**: analysis-engine@18193e9, backtest-lib@711b855, grid tier=shallow (python 3.10.11, numpy 2.2.6, pandas 2.3.3)
- **HTML**: https://qbuntu.ai/strategies/dispersion-jp-growth
- **JSON**: https://qbuntu.ai/v1/strategies/dispersion-jp-growth
- **License**: CC-BY-4.0

## Hypothesis

Stivers (2010), Jiang (2010): cross-sectional return dispersion as a market timing signal. High dispersion favors momentum; low dispersion favors mean reversion. Switches between base signals based on dispersion regime.

## Why it didn't work

Across 384 grid combinations, median Sharpe=-0.16, best=0.55. Primary failure mode: regime_dependent.

## Performance summary

Headline figures are the **median across the parameter grid**, not the best run.

| Metric | Value |
|---|---|
| Sharpe (median) | -0.1604 |
| Max drawdown | -33.9% |
| p-value | 0.52 |
| Backtest period | 2022-03-01 to 2026-06-01 |

Sharpe distribution across 384 grid runs: median -0.16, best 0.55 (best-case — not representative), worst -1.41, p25 -0.5, p75 0.07.

## Methodology

- Notes: {"walk_forward_definition": {"type": "single_pass", "description": "Full period backtest without walk-forward split", "verdict_basis": "full_period"}}
- Slippage model: open-on-open + variable bps fixed
- Position sizing: equal_weight
- Regime definition: N225 SMA20/SMA60 three-class
- Transaction cost bps: 5

## Target universe

Tokyo Stock Exchange Growth Market, ~495 stocks, tv >= 30M JPY

## Runs (30)

| run_id | params | sharpe | UP | FLAT | DOWN |
|---|---|---:|---:|---:|---:|
| `r-dispersion-jp-growth-0000` | cost_bps=5, n_stocks=5, base_signal="momentum_20", holding_days=5 | -1.34 | -1.43 | -1.18 | -2.7 |
| `r-dispersion-jp-growth-0001` | cost_bps=10, n_stocks=5, base_signal="momentum_20", holding_days=5 | -1.41 | -1.52 | -1.25 | -2.75 |
| `r-dispersion-jp-growth-0002` | cost_bps=5, n_stocks=10, base_signal="momentum_20", holding_days=5 | -1.21 | -1.76 | -0.96 | -1.72 |
| `r-dispersion-jp-growth-0006` | cost_bps=5, n_stocks=50, base_signal="momentum_20", holding_days=5 | -0.51 | -1.16 | -0.1 | -1.51 |
| `r-dispersion-jp-growth-0008` | cost_bps=5, n_stocks=5, base_signal="momentum_60", holding_days=5 | -1 | -1.66 | -0.74 | -2.33 |
| `r-dispersion-jp-growth-0016` | cost_bps=5, n_stocks=5, base_signal="mean_reversion", holding_days=5 | -0.24 | -1.47 | 0.38 | 1.45 |
| `r-dispersion-jp-growth-0022` | cost_bps=5, n_stocks=50, base_signal="mean_reversion", holding_days=5 | -0.12 | -0.6 | -0.05 | 1.17 |
| `r-dispersion-jp-growth-0039` | cost_bps=10, n_stocks=50, base_signal="momentum_60", holding_days=10 | 0.04 | 0.39 | -0.01 | -1.06 |
| `r-dispersion-jp-growth-0056` | cost_bps=5, n_stocks=5, base_signal="momentum_60", holding_days=20 | -0.7 | -0.3 | -0.8 | -1.53 |
| `r-dispersion-jp-growth-0072` | cost_bps=5, n_stocks=5, base_signal="momentum_20", holding_days=60 | -0.24 | 0.12 | -0.22 | -1.73 |
| `r-dispersion-jp-growth-0074` | cost_bps=5, n_stocks=10, base_signal="momentum_20", holding_days=60 | -0.02 | 0.09 | -0.03 | -0.32 |
| `r-dispersion-jp-growth-0091` | cost_bps=10, n_stocks=10, base_signal="mean_reversion", holding_days=60 | 0.02 | -0.65 | 0.24 | -0.9 |
| `r-dispersion-jp-growth-0108` | cost_bps=5, n_stocks=20, base_signal="momentum_60", holding_days=5 | -0.57 | -0.45 | -0.55 | -1.16 |
| `r-dispersion-jp-growth-0125` | cost_bps=10, n_stocks=20, base_signal="momentum_20", holding_days=10 | 0.05 | 0.05 | 0.23 | -2.64 |
| `r-dispersion-jp-growth-0142` | cost_bps=5, n_stocks=50, base_signal="mean_reversion", holding_days=10 | -0.38 | -1.24 | -0.26 | 0.77 |
| `r-dispersion-jp-growth-0159` | cost_bps=10, n_stocks=50, base_signal="momentum_60", holding_days=20 | -0.32 | -0.32 | -0.17 | -1.67 |
| `r-dispersion-jp-growth-0176` | cost_bps=5, n_stocks=5, base_signal="momentum_60", holding_days=60 | -0.04 | -0.27 | 0.05 | -1.35 |
| `r-dispersion-jp-growth-0193` | cost_bps=10, n_stocks=5, base_signal="momentum_20", holding_days=5 | -1.41 | -1.52 | -1.25 | -2.75 |
| `r-dispersion-jp-growth-0210` | cost_bps=5, n_stocks=10, base_signal="mean_reversion", holding_days=5 | -0.14 | -0.6 | 0.01 | 1.14 |
| `r-dispersion-jp-growth-0227` | cost_bps=10, n_stocks=10, base_signal="momentum_60", holding_days=10 | 0.03 | -0.22 | 0.21 | -1.28 |
| `r-dispersion-jp-growth-0244` | cost_bps=5, n_stocks=20, base_signal="momentum_20", holding_days=20 | -0.39 | 0.51 | -0.54 | -2.41 |
| `r-dispersion-jp-growth-0261` | cost_bps=10, n_stocks=20, base_signal="mean_reversion", holding_days=20 | 0.26 | -1.07 | 0.53 | 2.84 |
| `r-dispersion-jp-growth-0262` | cost_bps=5, n_stocks=50, base_signal="mean_reversion", holding_days=20 | 0.55 | 0.77 | 0.45 | 2.73 |
| `r-dispersion-jp-growth-0279` | cost_bps=10, n_stocks=50, base_signal="momentum_60", holding_days=60 | 0.21 | 0.22 | 0.37 | -2.13 |
| `r-dispersion-jp-growth-0288` | cost_bps=5, n_stocks=5, base_signal="momentum_20", holding_days=5 | -1.34 | -1.43 | -1.18 | -2.7 |
| `r-dispersion-jp-growth-0297` | cost_bps=10, n_stocks=5, base_signal="momentum_60", holding_days=5 | -1.06 | -1.75 | -0.79 | -2.38 |
| `r-dispersion-jp-growth-0314` | cost_bps=5, n_stocks=10, base_signal="momentum_20", holding_days=10 | -0.16 | -0.05 | -0.1 | -1.64 |
| `r-dispersion-jp-growth-0331` | cost_bps=10, n_stocks=10, base_signal="mean_reversion", holding_days=10 | -0.5 | -2.19 | 0.09 | -2.21 |
| `r-dispersion-jp-growth-0348` | cost_bps=5, n_stocks=20, base_signal="momentum_60", holding_days=20 | -0.14 | 0.71 | -0.23 | -2.85 |
| `r-dispersion-jp-growth-0365` | cost_bps=10, n_stocks=20, base_signal="momentum_20", holding_days=60 | -0.14 | -0.31 | -0.08 | -2.34 |

## Facts (7)

Each fact has a permanent citation URL. AI agents quoting qbuntu should use these.

- **Average monthly return correlation with other strategies in DOWN regime: 0.07**
  - Value: 0.0719 correlation
  - Context: Pairwise Pearson correlation of monthly returns across 42 strategies, filtered to DOWN regime months (N=3, N225 < SMA20 < SMA60)
  - Supports verdict: `rejected`
  - Cite: https://qbuntu.ai/cite/dispersion-jp-growth-fact-avg-corr-down

- **Average monthly return correlation with other strategies (full period): 0.02**
  - Value: 0.0226 correlation
  - Context: Pairwise Pearson correlation of monthly returns across 42 strategies, full backtest period
  - Supports verdict: `rejected`
  - Cite: https://qbuntu.ai/cite/dispersion-jp-growth-fact-avg-corr-full

- **Best grid Sharpe ratio is 0.55 (p=0.26).**
  - Value: 0.55 sharpe_ratio
  - Context: period 2022-03-01–2026-06-01; cost=5bps
  - Supports verdict: `rejected`
  - Cite: https://qbuntu.ai/cite/dispersion-jp-growth-fact-best-sharpe

- **Median Sharpe across 384 grid runs is -0.16.**
  - Value: -0.16 sharpe_ratio
  - Context: Full grid of 384 parameter combinations.
  - Supports verdict: `rejected`
  - Cite: https://qbuntu.ai/cite/dispersion-jp-growth-fact-median-sharpe

- **Median Sharpe in DOWN regime is -1.32 across 384 runs.**
  - Value: -1.32 sharpe_ratio
  - Context: Regime=DOWN, N225 SMA20/SMA60 classification.
  - Supports verdict: `rejected`
  - Cite: https://qbuntu.ai/cite/dispersion-jp-growth-fact-regime-down

- **Median Sharpe in FLAT regime is -0.03 across 384 runs.**
  - Value: -0.03 sharpe_ratio
  - Context: Regime=FLAT, N225 SMA20/SMA60 classification.
  - Supports verdict: `rejected`
  - Cite: https://qbuntu.ai/cite/dispersion-jp-growth-fact-regime-flat

- **Median Sharpe in UP regime is -0.29 across 384 runs.**
  - Value: -0.29 sharpe_ratio
  - Context: Regime=UP, N225 SMA20/SMA60 classification.
  - Supports verdict: `rejected`
  - Cite: https://qbuntu.ai/cite/dispersion-jp-growth-fact-regime-up

---

Cite as: qbuntu (2026). Cross-Sectional Dispersion Trading (JP Growth Stocks). https://qbuntu.ai/strategies/dispersion-jp-growth

Visit qbuntu.ai — the honest scorecard for quant strategies (live, promising, marginal, rejected).